Please specify model files.

[htw2012]

hello,I run the script, I find that there'e no corpus in it.Could you support me with some data ?

BTW,I use seq2seq that two problems I met:
1.I can save the model,but can't load the model.It occur some error like:

File "/usr/local/lib/python2.7/dist-packages/keras/utils/generic_utils.py", line 14, in get_from_module str(identifier)) Exception: Invalid layer: Seq2seq

2.AttentionSeq2Seq can't use.Could you run it?

Any help thanks for advance.

[galaxyh]

Hi, I am pretty sure there are problems in the code. Maybe I should close this repository. When I tested it, it just can't get translation right. I suggest you to look into Google TensorfFow example. There is a working implementation (Bengio, 2015) of machine translation. However, I didn't come up a good result using it to do English-Chinese translation, neither).

[htw2012]

Thank you your help.
Last time I try to use tensorflow in machine translation,but I can't run it owing to big corpus.
Could you do some toy examples like addition problem to validate the Attention implementation?